Abstract
For most actuarial modeling applications, model parameters are unknown and must be estimated. If the associated parameter estimation error is not recognized in the modeling, there is a good chance that a substantial portion of the adverse (and favorable) loss potential will appear to be diversified away in the aggregation process.
Keywords: modeling applications, parameters
Volume
Summer, Vol 2
Page
1-16
Year
2012
Categories
Actuarial Applications and Methodologies
Dynamic Risk Modeling
Publications
Casualty Actuarial Society E-Forum